freenode/#lisp - IRC Chatlog
Search
6:04:47
p0a
so far I wrote something using drakma and plump. Drakma retrieves the pages and plump is used to find the news titles
6:05:27
p0a
my issue is that reddit has everything in 1 line, making it hard to read the HTML to track down the relevant information. What should I do?
6:11:04
p0a
For example, I've noticed that the news articles have an attribute in <a> that other links don't have, but in general it is pretty cumbersome to extract exactly what I need every time
6:18:40
p0a
so it's pretty silly in what it does but the use is... you type (news) to get the frontpage news for the subreddits in the global var, and then you type (myread "worldnews") to read the titles of worldnews
6:25:14
MichaelRaskin
p0a: maybe look whether something like closure-css could be useful — you might be able to find someone who has written CSS selectors for that, and also Reddit's own CSS might be a hint. In general, data extraction from HTML has some amount of inherent (unless you control the other side…) complexity.
7:02:00
beizhia
well tbh, I dont really know where to begin. Should I make a separate asdf package for my tests? And do I need to symlink my project's dir in ~/common_lisp so asdf will find it?
7:03:13
beizhia
I think I might just be a bit spoiled by languages that set up the project structure for you, somewhat new to non-emacs lisp too. I just cant figure out what the normal thing to do it.
10:10:29
jmercouris
minion: memo for White_Flame: any examples of tree based GA in Lisp that you can point me to?
10:22:11
_death
in particular http://www.genetic-programming.org/gplittlelisp.html and https://jorgetavares.com/2017/05/03/gp-code-on-github/
11:09:13
jmercouris
_death: I saw the first source before but was put off from the lack of indentation